Using AI to Make Recruitment More Human (with Amber Wanner of Vette)

Paradox of the century? Hear how Vette uses AI to make recruitment more human, with instant interviews that make applicants feel they matter.
Daan van Rossum
Daan van Rossum
Founder & CEO, FlexOS
I founded FlexOS because I believe in a happier future of work. I write and host "Future Work," I'm a 2024 LinkedIn Top Voice, and was featured in the NYT, HBR, Economist, CNBC, Insider, and FastCo.
June 11, 2024
min read

🎧 Listen Now:

In today’s episode, we meet Amber Wanner, founder of Vette, who’s using AI to make recruiting more human.

We'll discuss how Amber’s methods provide instant, meaningful conversations, not just interviews, often with people in similar roles who may be suited better for that conversation than a recruiter. I love Amber’s mission of making people feel they matter and the unique ways they’ve done this.

Key Insights from Amber Wanner

My discussion with Amber unveiled several critical insights:

1. Making People Feel Valued

Amber said her purpose is to show as many people in the world that they matter. Recruitment can feel like a plight, but try to reenergize your process to emphasize human connection.

2. Instant Interviews

When you let people choose when to interview, as Vette does, you see that people choose to interview much faster than the usual process, and over 50% of the time outside of office hours. Does your current recruiting process support this?

3. From Interviews to Conversations

Amber is dedicated to transforming hiring from traditional interviews to conversations. She believes in allowing colleagues and individuals familiar with the role to interact with candidates rather than relying on recruiters with limited knowledge of the position. She said: 

“A warehouse worker talking to another worker flows better, allowing for a safer, more comfortable environment."

4. Using AI Purposefully

Amber found great ways to apply AI in places where it complements the human touch. For example, by matching candidates and interviewers like Uber does riders and drivers. They also use AI to summarize conversations and extract key data points. If you’re looking for ways to implement AI, this is a great place to start.

🔔 Available on:


Daan van Rossum: I came across through the people of the GPT community. I was putting together a top 40 of AI in HR companies, and then someone suggested looking into you. I checked it out and then realized that Hey, there's something really interesting here.

Can you share a little bit about vette, your platform? How does it work? What does it do, and where does it fit in? 

Amber Wanner: Yeah, absolutely. So, it's actually a gig platform. So we offer an applicant an interview as soon as they apply to a company's job. Right at that moment of intent, when they're expressing interest, we offer them the ability to speak to a human within seconds of them applying.

So we're working within the high volume, high turnover industry sector. And we're helping companies with retention and really understanding their applicants.

Daan van Rossum: How does that work? I saw a couple of screenshots, and that seems almost too magical that I apply for a job and I get a message and it says, Hey, someone is ready to interview you right now, which is about two to three weeks faster than any normal recruiting process. So how do you have people ready to interview someone? How do they know what questions to ask? How does it work on that side? 

Amber Wanner: So we are actually on a mission to change the industry from interviews to conversations. We found after launching the platform that our vets are gig workers and they are our neighbors; they are stay-at-home moms, retirees, and military veterans. They also could be the company's existing employees. So we can get to that in a little bit, but it's a conversation. It's a scripted conversation that we're able to really just understand the applicant, what matters to them, and the data takes care of the rest. So it's a data-driven conversation led by a human essentially.

Daan van Rossum: That's amazing. Amazing that you are able to also provide opportunities for those groups of people who may not be able to go into an office, as more companies are now moving back into the office. That sounds really interesting. So how does that scripted conversation work?

Do they get some prompts that they study, and is it dependent on the role that they're hiring for, meaning that you have to retrain those gig workers every time they start interviewing or start hiring for a new role? How's that work? 

Amber Wanner: Yeah. So the way the process works is that an applicant will apply to a company's job, whether it's through a job portal or through the ATS, and it's the moment that the applicant applies to the job that they get a text message saying, Hey, I saw you apply to this job. We have someone who can talk to you right now. Would you like to talk?

Daan van Rossum: And the candidate is like, what? Wait, wait, wait, what?

Amber Wanner: Originally, when I first started the company, I was like, Will applicants like to respond one and two? Will they want to interview right away? And so I started testing it out manually when I first started the company, and the responses started coming in, like wow, this is so fast. I can't believe how fast this is.

And on average, applicants are interviewing within a minute and 28 seconds of applying. We've actually eliminated the need for scheduling altogether. So they have the “interview now” link right in their text. So they can get to it whenever they're available. So that I'm measuring a heat map of the times of day that applicants want to interview.

But as soon as the applicant hits “interview now,” it searches who's online, just like Uber would, and it matches them up with a vet who's online. And the vetter accepts it, and the script pops right up on the vetter platform. And so, like I said, it's a conversation. So it's really what we really want to get to from changing it from interviews to conversations is kind of storytelling of the applicant and really getting to the meat of what the applicant truly needs to be happy and so companies can use that to retain them.

And so we found that they're just confiding in strangers. And so they're saying how much they actually really need and what the shift preference is. And so we're able to use that data to say, Okay, well, the company is not offering that, so it's not a fit for what the applicant needs. And so it's the data that's actually suggesting if the applicants are a fit or not.

Daan van Rossum: That's incredible. So intriguing. How do you find the people? How do you find the vets? Because you have to recruit those somehow as well. 

Amber Wanner: Yeah. So it's all been word of mouth. And originally, when I first started, I would recruit all of my Uber drivers. That was like, Hey, do you want to make some extra money in your free time? But no, as the platform started to advance and get more sophisticated, we started just having existing vetters refer other vetters. And so we're tracking the referral source, and it's all existing vets or heard about it from someone. And now we recently released something called that marketplace where companies could actually utilize their internal employees to be the vetters in their off hours.

Daan van Rossum: That makes total sense. So that's what I was wondering: is this, at the end of the day, really a platform play, or is it also something where you could just offer it up to companies again, really more from the software side? So, it sounds like you're betting on both sides, both the gig worker side and the platform software side.

Amber Wanner: Yeah. That's the cool part is that we're scaling humans, essentially like we're working with humans and utilizing technology to scale it. But what we started finding was that companies’ employees started applying to be vetters on our platform. And so we were like, This is interesting. And we started seeing that the data that we're collecting of what applicants are saying that they need to stay at a company or why they left the last company is because they don't have enough shifts and they can't pay their bills.

This is something that companies can offer them. To make some extra money in their free time by being a gig worker on the platform for their own company.

Daan van Rossum: Yeah, that's incredible. There really are a lot of analogies with Uber. You may be the first startup that actually is making Uber for X work. It actually works. 

Then, how do they capture the vetters? How do they capture the information that they're getting back? Because obviously, there's a lot of skill in doing interviews really well. So how do you train them for that? And then how do they process the information that they're getting, or is that just recorded and it's done by the system on the backend?

Amber Wanner: Yeah. So that's actually where we use AI. So we're taking those conversations, and we're summarizing the conversation. We're pulling out data from it. So really the vetter’s purpose is just to show this person that they matter to really just have that one voice on the other line, showing this person who maybe is scared that they may not get a job or it's giving them hope.

And so the better purpose is just to be that warm voice where the conversation we're pulling in data and information out of the conversation along the way. So we are using AI to summarize it. We're really big on words. So if someone's saying, I love or I'm passionate about, we can pull out what this person really enjoys doing and do sentiment analysis there as well.

Daan van Rossum: That's so interesting. But there must be some people who are really good at that. If you think about it on the vetters side, there must be some people who get a lot out of the candidate and maybe some people who are, like, slightly less versed. Do you, over time, also optimize for who is getting those really great outputs versus maybe people who just don't really fit with the platform? Just like on Uber, you would have some rating system. There was some feedback loop on who is right. And then do you make decisions based on that? 

Amber Wanner: Yes, we do have a rating system. So after the conversation is done with the applicant, the applicant gets a text saying, asking how their experience was. So they're able to rate them. But what we actually found is that it's really interesting when say a warehouse worker is talking to another warehouse worker or someone that they can relate to; the conversation flows better rather than it being someone who's a corporate person talking to a warehouse worker. So we find that the vetters that can relate to the applicants are really able to pull more out of the conversation because the applicant feels more safe and more comfortable. So it's almost like an equilibrium, the applicant feels with this matter that is similar to them.

Daan van Rossum: This may just be one of the most interesting uses of technology plus humans that I have heard. Because I can totally imagine that if you speak to someone who has done that job or is doing that job, the conversation is totally different and you can actually get into the details, which is typically what a recruiter wouldn't be able to do. They wouldn't really know in depth what the role is like. So you wouldn't have that same sense of connection. That's super interesting. 

How's the candidate's feedback? How do they feel when they connect with someone like that? Because again, that interview experience compared to what they would be experiencing with other companies, where there's a more traditional recruiting process, would be night and day. 

Amber Wanner: Yeah. So the applicants love it for a couple of different reasons. One, they can interview sometimes at 1 in the morning when they're getting off their shift. 24x7, they're able to talk to someone. And so the feedback is very positive that they don't need to schedule. They can do it when they're ready. They literally just hit a button, and they're able to speak to a human. This also allows for the companies to capture the applicants as soon as again, are okay, I'm ready to look for a new job.

So whoever gets that applicant first shows them that they matter. They're the ones that are going to get the applicant. And so the second thing is that a while back I was talking to an investor, and they said to me, What are you going to do when AI comes out and starts sounding like a human? Now we know GPT-4 recently came out, and I'm not gonna lie. I stayed up for 4 nights researching like the human brain and like what the difference is between AI. And I was listening and studying these conversations because we record all the conversations.

We started finding that humans have one, a why, but also humans have memories that they can then tap into. They have relatability that allows a person to feel safe and comfortable, and trust is built. Whereas with AI, I'll give you an example. There is someone who was from. New Jersey, and they were like, Oh, did you grow up going to the Bridgewater Commons mall? I went there with my grandparents growing up.

That relatability allows the person; it's almost like the rapport is built. Whereas with AI, and I listened to the GPT-4, and it can be funny and everything like that. But the moment that the warehouse worker or the security guard realizes that they're not talking to a human or that it just isn't the same emotional connection that they have with talking to a real human.

Daan van Rossum: As you said, it's exactly that point that even AI could mimic and could say all these facts about New Jersey, but the person would know that they didn't grow up there. They would know that they didn't have grandparents who are from there, and so that's obviously the big difference.

A real human is a real human and will always be a real human. Probably the one place where it may be worth still investing in humans is in the realm of where it's really about that human connection, where there's actually a premium for being with that human. 

That obviously leads to the question, Where does this sit in the landscape? Because you do now have platforms out there. And I've spoken to Sapia and Eightfold on this season. And then just the interview with David from PeopleGPT Juicebox. There's a lot of platforms out there that, all in their own way, bring AI to recruiting. And some of that is much more in an automated way where no human is involved.

And therefore it scales infinitely. So when you go and pitch this, are you being compared against those solutions? There must be a higher cost. So how do you justify that? And then how do you also keep the customers happy as they start using the platform? 

Amber Wanner: Yeah, so we really want to make sure that we're narrowing the funnel of applicants that actually get to the conversation. We don't want to waste the applicant's time. We don't want to waste the company's time or money. And so we do use AI in the SMS communication. So if someone asks a question like, What is the hourly rate?" they can almost self-disqualify themselves in that SMS. They also have knockouts that companies could add before they speak to the human.

As a business, our goal is not to necessarily get more conversations. It's the right conversation. It's the right applicants through the funnel to get the conversations that are closest to the fit. And so what we're also doing is we're taking the interview data, and we're saying, Okay, these are all the things that match up with what you express that you're looking for.

And so we can get more accurate with who is actually getting to the company, and so fixed cost is obviously paying the vetters. So we have ways that companies could scale up and down based on their needs and their costs. So, it doesn't necessarily cost more. We did find that companies are having to interview fewer applicants per hire because they're getting the right one.

Daan van Rossum: Compared to traditional recruiting, for sure, it would be more affordable. It could definitely be more affordable, but going against a complete total AI platform, there would be a premium. And I'm just curious about that because you were just mentioning how you're really looking at this as a solution for high-hiring, high-turnover roles. Is that the role that company wants to play in that investment? 

Amber Wanner: Yeah. We are a company with 92% of their recruiting and tech costs. We have packages that start off with basic, but it all is based on the engagements of applicants and whether it's the touch points of the SMS, whether it's a knockout or the conversation. It really comes down to getting that quality applicant in the funnel.

But it starts at a $60,000 platform fee, and then there's different tiers from there.

Daan van Rossum: Super interesting. I'm just really thinking about where that sits compared to other platforms and the traditional approach, and I'm sure you have to explain a lot because it's still pretty early days for this approach.

What is the conversation like when you're trying to sell this to a company? Are people ready for it? Is there still a lot of education required? How does that work? 

Amber Wanner: So we do have a free trial that companies are able to test it out to see how it works initially. But what we found was that when they start seeing the results, we have an automation into any ATS pretty much from the top of the funnel.

As soon as the applicant applies, we're able to actually put data back into the ATS, update the status, and move it through the hiring process. So, it works within the ATSs, but we have found that some companies have been able to reduce their ad spend and some line items by using that.

Daan van Rossum: They can do self-sign up. They can try it out, and they can see the benefits, and then start really using the platform. One other thing you mentioned is that applicants really like it because they can talk to people whenever they are ready, which I think is again a great service to the applicants.

It's such a great first impression. If you're applying for a role, comparing that to traditional recruiting, I think that's like night and day. Are you then seeing that people are actually having those conversations at very different times than what companies normally would think, because normally you're just trying to fit them into a typical 9-to-5 schedule with maybe the groups that you mentioned that may not be the right time for them? What's that data telling you? 

Amber Wanner: Yeah. So 48.5% of applicants are interviewing off hours. So outside of the, almost half, Monday through Friday, 9 to 5. So between weekends and nights, like I said, we have some interviews going off, 9 p.m., 1 a.m., 5 a.m. It ranges all over, and because our vets are located across the US and eventually across the world, they could also request their conversation in a different language too. We're able to meet the applicant right where they are.

And like I said, I think that one of the biggest advantages is that when you really think about it, what are the humans reacting to? So if it's AI alone, there's really not a whole lot of accountability for that applicant to show up, whereas when they talk to humans, they feel more emotionally connected.

Daan van Rossum: It's super interesting. I think the really interesting thing here is that. It just turns the whole thing on its head. I think this is as I'm listening to you; it just turns the whole thing on its head. So normally the candidate really fits what the company wants, in terms of when to interview and in which language and how. And now that you're giving people choice, like the data is actually telling you that for half of the people, they would rather speak at a time outside of normal office hours.

So you just see how the current process actually is completely not fitting with the expectations, and then again, I'm just really thinking about that experience because everyone applies for multiple roles at the same time. If you have that experience with one company versus another company where you just go into the maybe two to three weeks before someone reaches out and then you're trying to do the scheduling game and all that, just as a candidate experience and as an employer brand initiative and as what I would tell other people about that company, that just seems like an incredible competitive advantage as a company.

Amber Wanner: Yeah. And not only that, again, we're talking about hourly workers. So these are people who are scared that they're not going to be able to afford their rent this month if they don't get a job. And so time literally is of the essence for them to be able to move forward in the process in order to provide for themselves.

That's why, with the data, when we're pulling, we ask a question in the conversation that says, Why did you leave your last job, and what could they have done to make you stay, and what do you need in the next company? And the two most common responses are, one, they don't have enough shifts at the last job. And so therefore they can't pay their bills.

So those are the two main factors from which we know our data. We know why applicants are leaving companies jobs. Like these applicants are telling us straight from their mouth, why they're leaving, these companies, these large organizations. We have all the data of why these people are leaving and because they can't pay their bills or their basic needs.

The cost of living is really high. At $20 an hour, it's really sad, but that's the reality, and the companies that are now with our platform are creating gigs, working inside of their organization where they could use a warehouse worker to interview an applicant applying for a warehouse job and provide an additional source of income for them. These companies are shown that they care, and it's putting money back into their employees, essentially

Daan van Rossum: It seems like a win-win. Any downsides to this platform, this approach?

Amber Wanner: No, I don't know if you're familiar with something called Pareto improvement, but companies right now are trying to create efficiencies. But that means that someone is going to lose something. And so with Pareto improvement, it's utilizing your existing resources to grow your business. And that's what this platform offers these companies today.

Daan van Rossum: Another really hot topic in this space is bias, especially when you're working with AI, and when you're working with AI plus humans, there's always a risk that bias creeps in and how the model that you're using is trains or how certain selections are made by the AI. How do you make sure that, in that process of trying to get the right candidates to the right role, that bias doesn't creep in or as little as possible?

Amber Wanner: We're relying on job descriptions that literally are from, I don't know, the 90s. And so we're having a hiring manager go through that, and we're taking what the hiring manager is saying that they can offer and we're matching it to what the applicant is saying that they need.

And so we're able to use real data to say, Okay, this person needs this, and this is what the company is offering. If it's not aligned, it's not a fit. So there's no bias in the sense that it's like, Oh, I don't know if you'd be right for the job. It's literally and again, we're talking about hourly jobs like these high-turnover, high-volume jobs that companies are hiring for.

Why is it high turnover? It is high turnover because we're not listening to what the applicant needs or the hiring manager. It's not just high turnover among the applicants. It's also high turnover in hiring managers, the line managers, and the operations people on the floor.

And so when both parties feel heard and we can align what they need versus what they're offering, then that's where there isn't any bias. It's literally based on data and what this person said straight from their mouth.

Because when someone is interviewing with a recruiter, they're interviewing with someone at the company, and they tend to say what they think they want to hear. And that's really where that is really cool because it's talking to a stranger. It's like telling this person, Okay, this is what I need. And it's a conversation. It's not an interview. It's not a, tell me a time where you had to do blah, blah, blah.

Again, it's storytelling. It's the conversation. It is consistent, and it's storytelling. So what did you like to do as a kid when you're younger? What made you happy? And you can start to understand things like what this person really enjoys, leading up to where they are today, and where they want to go. You get a really lot of information when you ask a question and then you just listen.

Daan van Rossum: Again, especially when it's someone that you feel that immediate connection with because they come in on your schedule and they speak your language, and they may have a background in that role.

There's just much more of a connection there. I really sense from the way you're speaking about this that you're very passionate, especially about that human side. Maybe you can share a little bit as we're getting to the end about why you started this business and what keeps you going?

Amber Wanner: My purpose for this company is a couple of things. One is to show as many people in the world that they matter as possible. Just so happen to be doing it through interviews and helping businesses grow with this platform, but I want to one day go into another country where people are slaving away at, say, a call center or whatever and say, You don't need to be there doing that. You can go home and be with your family; go online when you want to work offline when you don't.

It is gig work for tasks that businesses are outsourcing right now. We've started in interviews, but we have companies that are using that for other things, like after-hours answering and really utilizing the vetters to perform these tasks that they were otherwise outsourcing.

To really provide options for people in even marginalized communities. I want this platform to live far beyond me. I know it sounds crazy, but this is bigger than myself. And I want people who are stay-at-home moms and people who maybe didn't have the easiest life to have options to be able to provide for themselves.

And that's really the motivation. And I listened to these vets because, again, they're all recorded. And I hear what happens in these calls. The hope that these people have when talking to another human being.

Daan van Rossum: That's beautiful. Thanks so much for sharing the story. Thanks so much for being on. Really appreciate it. I loved learning more about it. And especially now hearing at the end that mission behind it and making people feel that they matter, which I think is very lacking in recruiting and generally in life. It's really amazing. So thanks so much for sharing.

Amber Wanner: Thank you for having me. I really appreciate it.

You Might Also Like …

All articles about

Future Work

A weekly column and podcast on the remote, hybrid, and AI-driven future of work. By FlexOS founder Daan van Rossum.